Multi-Strategy Guided Diffusion via Sparse Masking Temporal Reweighting Distribution Correction

2025-09-07

AI Summary

To address severe projection missing and challenges in global structural modeling for sparse-view CT reconstruction, this paper proposes STRIDE—a sparse-condition-guided time-varying diffusion model. Methodologically, STRIDE introduces a novel time-dependent sparse-condition reweighting mechanism that dynamically integrates prior information during denoising to accurately recover missing projections. It employs a multi-band dual-branch network to jointly optimize low-frequency structural fidelity and high-frequency detail preservation, augmented by linear regression-based distribution calibration to enhance generation consistency. Evaluated on multiple public and real-world datasets, STRIDE consistently outperforms state-of-the-art methods: PSNR improves by up to 2.58 dB, SSIM by 2.37%, and MSE decreases by 0.236. Moreover, it demonstrates superior robustness and generalizability in structural preservation, fine-detail recovery, and artifact suppression.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Structured Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Bridging structured and unstructured data

Abstract

Diffusion models have demonstrated remarkable generative capabilities in image processing tasks. We propose a Sparse condition Temporal Rewighted Integrated Distribution Estimation guided diffusion model (STRIDE) for sparse-view CT reconstruction. Specifically, we design a joint training mechanism guided by sparse conditional probabilities to facilitate the model effective learning of missing projection view completion and global information modeling. Based on systematic theoretical analysis, we propose a temporally varying sparse condition reweighting guidance strategy to dynamically adjusts weights during the progressive denoising process from pure noise to the real image, enabling the model to progressively perceive sparse-view information. The linear regression is employed to correct distributional shifts between known and generated data, mitigating inconsistencies arising during the guidance process. Furthermore, we construct a dual-network parallel architecture to perform global correction and optimization across multiple sub-frequency components, thereby effectively improving the model capability in both detail restoration and structural preservation, ultimately achieving high-quality image reconstruction. Experimental results on both public and real datasets demonstrate that the proposed method achieves the best improvement of 2.58 dB in PSNR, increase of 2.37% in SSIM, and reduction of 0.236 in MSE compared to the best-performing baseline methods. The reconstructed images exhibit excellent generalization and robustness in terms of structural consistency, detail restoration, and artifact suppression.

Problem

• Sparse-view CT reconstruction using diffusion models
• Dynamic sparse condition reweighting during denoising process
• Correcting distribution shifts between known and generated data

Innovation

• Sparse conditional probability guided joint training
• Temporal reweighting strategy for dynamic denoising adjustment
• Dual-network parallel architecture for frequency component optimization